A small project based on an assignment, see Assignment.pdf. See [Project Info](Project Info.pdf) for more information about the assignment itself.
The goal of the assignment is to study data from a trucking company. A database of every Routes taken by a driver as well as every individual step of this route, from one town to another, is provided to us. However, it is completely unsorted, and it is up to us to process it to extract useful information from it. Of course, once we're done, we need to display this information to the user, which we do using GNU Plot.
The format of the database is a simple CSV file, formatted as such :
| Route ID | Step ID | Town A | Town B | Distance | Driver |
|---|---|---|---|---|---|
| 123507 | 2 | BEYCHAC ET CAILLAU | LATRESNE | 13.617 | Laura NAVARRO |
| 107421 | 6 | MARIGNIER | FAUCIGNY | 9.953 | Eva WEBER |
| 39435 | 18 | ECLANCE | ROLAMPONT | 63.44 | Evgeniy SMITH |
| 42199 | 18 | CAUBON ST SAUVEUR | VAUSSEROUX | 219.856 | Tatiana RODRIGUEZ |
| 116783 | 2 | GIRAUVOISIN | BITSCHWILLER LES THANN | 152.254 | Anna KUZNETSOVA |
example from the first few lines of the CSV file provided to us as the file that our project must be able to process
cy-trucks DATA_LOCATION ARG
You may have to specify bash cy-trucks DATA_LOCATION ARG if the script isn't executable on your machine.
where DATA_LOCATION is the path to your CSV file (will be copied under data/data.csv) and ARG is either -d1, -d2, -l, -s or -t. See -h or man ./help for more information.
Go to the progc/ directory, and simply run make. You may run make clean to remove the build binary. The binary will be dropped in progc/bin under the name cy-trucks.exe. The .exe extension is to mainatin WSL compatibility, it is perfectly runnable under any Linux-based system.
cy-trucks.exe DATA_LOCATION ARG
where DATA_LOCATION is the path to your CSV file and ARG is either -s or -t.
[fleefie@providence Projet-Info-S3]$ bash main.sh data.csv -d1
Compiled C code
Starting D1 process..
Program ran in 3.518 seconds.
[fleefie@providence Projet-Info-S3]$ bash main.sh data.csv -d2
Compiled C code
Starting D2 process..
Program ran in 3.906 seconds.
[fleefie@providence Projet-Info-S3]$ bash main.sh data.csv -l
Compiled C code
Starting L process..
Program ran in 5.196 seconds.
[fleefie@providence Projet-Info-S3]$ bash main.sh data.csv -s
Compiled C code
Starting S process..
Program ran in 4.859 seconds.
[fleefie@providence Projet-Info-S3]$ bash main.sh data.csv -t
Compiled C code
Starting T process..
Expect this one to take a while !
Program ran in 17.757 seconds.
Top 10 towns by number of routes passing by them, sorted by alphabetical order, in descending order.
Tools used : C code, see progc/
Graph of the top 50 of routes by the largest and smallest distance of their steps, as well as the average distance of a route's step. Sorted by the difference between the largest and smallest steps of a route, in descending order.
Tools used : C code, see progc/
This project depends on BASh, Make, the GNU standard library, and either ImageMagick or FFMPEG. Both of the latter are installed in common Linux distributions, but as a failsafe, if they aren't found, the resulting images of -d1 and -d2 will simply not be rotated. This is done using BASh's hash command, so no need to worry about needing to fix the code yourself if you somehow don't have either of those dependencies.
See : GNUPlot scripts README. While not mentioned in the main.sh file, GNU.org for the help with SEd regex. As linked in the comments of main.sh, the GNU online software manual also helped us greatly with the use of gAWK. ImageMagick's convert command as well as the filters from FFMPEG were both implemented as means to rotate our plot's images, so credit is due where it is due.
So far, no bugs were found in our current code.
- Ask the user how many data points they would like to output in their plots. This would require advanced scripting, and while it feels like it could be a part of the scope of this project, our code uses too many crutches that depend on the data generation being relatively simple.
- Replacing most instances of sorting through arrays by balanced BSTs that allow duplicate values. Our BSTs not working with dupes is why we had to use arrays for some sorts and why we needed to do a sub-BST for every node of a BST in the T process, leading to it being as long as it is. This enhancement would allow us to greatly increase the scope of this project, such as permitting the implementation of the above feature.